One of the most unusual exoplanet systems discovered so far is WASP-47.  The first planet discovered in this system was a seemingly garden-variety hot Jupiter, WASP-47 b.  WASP-47 b is the size and mass of Jupiter and orbits its star in just 4 days.  It was discovered in transit, meaning that it passes in front of its star and blocks a small amount of starlight every orbit.  Although the origins of hot Jupiters are mysterious, they inhabit about 1% of solar systems and are not particularly rare among the 4000+ currently catalogued exoplanets.

Several years later, NASA’s reanimated Kepler Space Telescope K2 pointed its gaze in the direction of WASP-47, and what K2 found shocked the exoplanet community.  All previous studies of hot Jupiters have found that these are isolated planets, without a single other planet in the system.  Some of the most popular theories suggest that hot Jupiters, on their journey from where they form toward the star, violently eject other planets from their orbits.  WASP-47 b, however, has two neighboring planets, both smaller than Neptune, that peacefully coexist with the gentle giant.  WASP-47 d is 2.3 times the diameter of Earth and orbits the star in 9 days.  On the other side of the giant planet, WASP-47 e is 1.7 times the diameter of Earth and orbits the star in just 20 hours.  Like the hot Jupiter, the small planets transit their star, but Earth’s turbulent atmosphere had hidden them from the original ground-based transit survey.  WASP-47 is the only example of a hot Jupiter with nearby neighbors.

A cartoon of the WASP-47 system (not to scale). Left to right: the star, a super-Earth sized planet at 0.8 days (planet “e”), a hot Jupiter at 4.2 days (planet “b”), a sub-Neptune sized planet at 9.0 days (planet “d”), and a non-transiting Jupiter-mass planet at about 600 days (planet “c”). The star has a sun-like mass, radius, and temperature, but has a much greater iron abundance than the sun. The spin angle of the sun is more or less aligned with the orbits of the transiting planets.  The inner three planets have orbits that are more circular than most of the solar system planets.  The outer planet (c) has a significantly elliptical orbit, similar to that of Pluto.  The two giant planets are mostly gas.  The sub-Neptune (d) has a significant gaseous layer as well.  The gas content of the super-Earth (e) is not well determined.

How did the hot Jupiter and its companions arrive at their current orbits without destruction?  Enriching the trail of clues is WASP-47 c, a Jupiter-mass or heavier planet with a 600 day, elliptical orbit that was discovered by measuring the stellar wobble.  Could the distant planet have caused the inward migration of the hot Jupiter, and if so, how did the small planets survive the journey?

The orbital properties and compositions of the WASP-47 planets have left us with clues about the formation story of this system.  In a recent paper I led, my teammates and I measured precise masses and orbital properties of all the planets in WASP-47.  A summary of our measurements, along with measurements from previous scientific work, is shown in the illustration of the system above.

We found that the hot Jupiter and the distant giant planet are both Jupiter-mass, and they are mostly composed of hydrogen gas.  Although hydrogen is rare on rocky planets like Earth, hydrogen is the most common element in stars and is readily available in the gas disk that precedes planet formation.  Our own solar system gas giants Jupiter and Saturn are also rich in hydrogen.  Like the solar system giants, the two giant planets of WASP-47 must have formed early in the system’s history, before the gas in the disk blew away.

On the other hand, the densities and sizes of the small planets indicate that these planets are gas-poor.  We found that the innermost planet “e” (1.7 Earth radii, 9 Earth masses) is consistent with a bare rocky composition, and the other small planet “d” (3.6 Earth radii, 14 Earth masses) likely has about 1% of its mass in a gaseous envelope.  Planets like these can form later in the system’s history, when the gas disk is almost completely gone (but enough gas needs to remain to form the gaseous envelope on planet d).

One way to explain the unusual architecture of the WASP-47 system is to invoke a mechanism wherein the giant planets form early, and the small planets form later.  If the hot Jupiter can form, move, and finally park in its current location before the gas disk dissipates, the small planets can form at their current orbits without disruption.  This is called two-stage planet formation.  Since our own solar system likely formed in at least two stages, it is reasonable to call upon such a mechanism to explain WASP-47.  A cartoon of some possible two-stage planet formation pathways for WASP-47 is shown below.

This diagram shows how planet-forming material might have moved in the WASP-47 system over time. The small planets could have formed as late as the era in which the gas disk was dissipating (halfway up the time axis), but the gas-rich giant planets must have formed earlier in order to accumulate their gas. A variety of theories can describe the early accumulation and motion of the material that eventually formed the giant planets. There are two key dynamical clues about the histories of the giant planets. 1) The hot Jupiter (b) must have parked with very low eccentricity before the small planets formed; otherwise the small planets would not be there. 2) The cold giant planet must have acquired its eccentric orbit after the gas disk dissipated; otherwise, the gas disk would have circularized its orbit.

Now that we have outlined some plausible rules governing the formation of the three inner planets, what can we deduce about the long-period giant?  The eccentricity (deviation from circular shape) of the distant planet’s orbit offers some insight.  Planets are thought to form in more-or-less circular orbits, but their eccentricities can increase when the planets exchange angular momentum with their surroundings.  While the gas disk is present, the eccentricities of planets can only rise to about 0.1 (i.e. 10% deviation from a circular orbit) before the gas disk counteracts their deviant tendencies.  However, the eccentricity of WASP-47 c is 0.28, meaning that it acquired its eccentric orbit after the gas disk dissipated.  Could the distant giant have exchanged angular momentum with the present-day hot Jupiter?  Although the hot Jupiter is massive enough to have survived an exchange of so much angular momentum, we just concluded that the hot Jupiter likely parked in its current close-in orbit–and with zero eccentricity–before the gas disk dissipated, in order to allow the small planets to form.  Since the hot Jupiter was not in the right place at the right time to give the distant giant planet a kick, somebody else must have done it.  This suggests that another giant planet might still reside somewhere in the WASP-47 system, waiting to be discovered.  Alternatively, such a planet might have been ejected during the angular momentum exchange, or the momentum swap might have been caused by a passing star long ago.

Regardless of who swapped angular momentum with the distant giant, our census of the WASP-47 system is far from complete.  Future studies will strive to reveal the presence of additional planets in the system, and also to characterize the atmospheres and interior physical properties of the known planets.  Identifying the present-day compositions of the planets will give us additional clues about where and how these planets formed, and will perhaps clarify why systems like WASP-47 are so rare.